本文選題：土壤呼吸　切入點(diǎn)：施氮　出處：《中南林業(yè)科技大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

【摘要】：目前施氮對(duì)土壤呼吸影響的研究大多基于實(shí)驗(yàn)觀測(cè)結(jié)果,受試驗(yàn)地自然條件的限制,不能研究在一定條件范圍內(nèi)土壤呼吸對(duì)施氮響應(yīng)的連續(xù)變化過程。通過噴灑NH4NO3水溶液,設(shè)置對(duì)照(C,no N added),低氮(L,5gNm-2a-1),中氮(M,15gNm-2a-1),高氮(H,30gNm-2a-1)4種處理水平,使用GA-BP(人工神經(jīng)網(wǎng)絡(luò))建立樟樹林土壤呼吸對(duì)施氮響應(yīng)的模型,通過模型模擬揭示了土壤呼吸對(duì)施氮響應(yīng)的變化。結(jié)果如下:(1)GA-BP人工神經(jīng)網(wǎng)絡(luò)可以較好的模擬亞熱帶樟樹林土壤呼吸在不同施氮量條件下的響應(yīng)過程。通過模型仿真得到的結(jié)果連續(xù)、完整,解釋了0.70-0.74的施氮水平、土壤溫度、土壤濕度交互作用下的土壤呼吸過程。(2)繪制了 土壤呼吸與土壤濕度和土壤溫度的響應(yīng)曲面圖及對(duì)應(yīng)的等高線圖。通過響應(yīng)曲面圖和變動(dòng)率圖直觀的描述了不同施氮水平下土壤呼吸關(guān)于土壤溫濕度的連續(xù)完整的變化過程。施氮在總體上抑制了土壤呼吸速率,但在高氮處理下土壤呼吸速率較低氮與中氮處理出現(xiàn)了增加,但未達(dá)到對(duì)照處理的水平。較低的土壤呼吸速率主要出現(xiàn)在土壤溫度降較低區(qū)域,隨著施氮量變化出現(xiàn)在不同的土壤濕度上;而較高的土壤呼吸速率主要出現(xiàn)在土壤濕度較高的區(qū)域,并隨著施氮量的變化出現(xiàn)在不同的土壤溫度上。(3)土壤呼吸總速率、低土壤濕度范圍的土壤呼吸速率和高土壤濕度范圍的土壤呼吸速率,三者比例從對(duì)照處理的1:0.83:1.29,到低氮處理的1:0.84:1.18,再至中氮處理的1:0.95:1.08,最后到高氮處理的1:1.04:0.96,隨著施氮量的增加,不同濕度范圍內(nèi)土壤呼吸有向均值趨近的現(xiàn)象。(4)利用GA-BP人工神經(jīng)網(wǎng)絡(luò)特點(diǎn),結(jié)合輸入層-隱含層-輸出層之間的權(quán)值,通過Garson算法計(jì)算求出影響土壤呼吸各因素的程度。從計(jì)算結(jié)果中可以發(fā)現(xiàn),凋落物水平對(duì)土壤呼吸的影響貢獻(xiàn)率最大,其余三個(gè)因素對(duì)土壤呼吸貢獻(xiàn)率基本一致。
[Abstract]:At present, most of the studies on the effects of nitrogen application on soil respiration are based on the results of experimental observation. Limited by the natural conditions of the experimental site, the continuous variation process of soil respiration response to nitrogen application in a certain range of conditions can not be studied. NH4NO3 aqueous solution is sprayed. The response of soil respiration to nitrogen application in camphor forest was established by using GA-BP- (artificial neural network) at four treatment levels, namely, no N added, low nitrogen (Ln) 5g Nm-2a-1, medium N ~ (2 +) ~ (15) g Nm ~ (-2) -1a ~ (-1), high N ~ (2 +) H ~ (2 +) ~ (30) Nm ~ (-2) ~ (-1) ~ (-1). The response of soil respiration to nitrogen application was revealed by model simulation. The results are as follows: 1: 1 + GA-BP artificial neural network can better simulate the response process of soil respiration in subtropical camphor forest under different nitrogen application rates. The true result is continuous, Complete, explaining the N application level of 0.70-0.74, soil temperature, Soil respiration process under the interaction of soil moisture. 2) the response surface map of soil respiration, soil moisture and soil temperature and the corresponding contour map are drawn. The different responses are intuitively described by the response surface map and the change rate map. The continuous and complete variation of soil respiration in soil temperature and humidity at nitrogen application level. However, the soil respiration rate increased under high nitrogen treatment, but it did not reach the level of the control treatment. The lower soil respiration rate mainly occurred in the lower soil temperature drop area, but it did not reach the level of the control treatment, but the soil respiration rate increased under the condition of high nitrogen treatment, but it did not reach the control level. The higher soil respiration rate appeared in the higher soil moisture area, and the higher nitrogen application rate appeared in the different soil temperature, and the higher soil respiration rate appeared in different soil temperature, but the higher soil respiration rate appeared in the higher soil moisture area, and the higher soil respiration rate appeared in the different soil temperature with the change of nitrogen application rate, and the higher soil respiration rate appeared in the area with higher soil moisture content. The soil respiration rates in low soil moisture range and high soil moisture range ranged from 1: 0.83: 1.29 in control, 1: 0.84: 1.18 to 1: 0.84: 1.18 in low nitrogen treatment, 1: 0.95 to 1.08 in medium nitrogen treatment, and finally to 11.0: 4: 0.96 in high nitrogen treatment. With the increase of nitrogen application rate, In the range of different moisture, the phenomenon that the average value of soil respiration approaches to the average value. (4) using the characteristics of GA-BP artificial neural network, combining the weights between input layer, hidden layer and output layer, The degree of influencing factors of soil respiration was calculated by Garson algorithm. The results showed that litter level had the largest contribution to soil respiration, while the other three factors had the same contribution to soil respiration.
【學(xué)位授予單位】：中南林業(yè)科技大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：S714
，

本文編號(hào)：1579669